摘要
目的 应用流行病学及临床资料构建肺癌风险预测模型。方法 回顾性分析344例肺部病变患者的人口学特征、临床资料及实验室检查资料等,应用LASSO回归筛选变量,构建肺癌风险预测模型并评估模型预测效果。结果 性别(OR=0.242,95%CI:1.305~3.928,P=0.004)、咳嗽(OR=2.028,95%CI:1.167~3.531,P=0.012)、吸烟(OR=3.198,95%CI:1.744~6.136,P<0.001)、慢阻肺1级(OR=0.249,95%CI:0.127~0.474,P<0.001)和慢阻肺2级(OR=0.449,95%CI:0.233~0.846,P=0.015)是肺癌发病的独立危险因素;受试者工作曲线下面积为0.731,95%CI:0.672~0.789。校准曲线分析结果显示该模型有较好的预测性能。结论 通过临床基础资料构建的肺癌风险预测模型,对提高肺癌早期诊治具有潜在的价值。
Objective This paper aims to construct a risk prediction model of lung cancer by using epidemiological and clinical data. Methods The demographic characteristics, clinical data and laboratory examination of 344 patients with pulmonary lesions were analyzed retrospectively. LASSO regression was used to screen variables, construct lung cancer risk prediction model and evaluate the prediction effect of the model. Results Gender(OR=0.242, 95%CI:1.305-3.928, P=0.004), cough(OR=2.028, 95%CI:1.167-3.531, P=0.012), smoking(OR=3.198, 95%CI:1.744-6.136, P<0.001), chronic obstructive pulmonary disease(COPD) grade 1(OR=0.249, 95%CI:0.127-0.474, P<0.001) and COPD grade 2(OR=0.449, 95%CI:0.233-0.846, P=0.015) were independent risk factors for lung cancer. The area under receiver operating characteristic curve was 0.731(95%CI: 0.672-0.789). The results of calibration curve analysis show that the model has good prediction performance. Conclusion The lung cancer risk prediction model based on clinical basic data has potential value in improving the early diagnosis and treatment of lung cancer.
作者
刘姿
王金来
贾建超
张雷明
LIU Zi;WANG Jin-lai;JIA Jian-chao;ZHANG Lei-ming(Department of Respiratory and Critical Care Medicine,Henan Provincial People's Hospital,Zhengzhou University People's Hospital,Henan University People's Hospital,Zhengzhou 450003,China;不详)
出处
《中国卫生检验杂志》
CAS
2022年第24期3007-3010,共4页
Chinese Journal of Health Laboratory Technology
基金
2021年度河南省档案科技项目计划(2021-X-11)
2021年度河南省档案科技项目计划(2021-X-12)。
关键词
肺癌
预测模型
LASSO回归
Lung cancer
Prediction model
LASSO regression
